![Sample our Engineering & Technology journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5933/TOC-Subject-Sample-2021-Engineering-Tech.jpg"})
Abstract
Monitoring health conditions over a human body to detect anomalies is a multidisciplinary task, which involves anatomy, artificial intelligence, and sensing and computing networks. A wearable wireless sensor network (WWSN) turns into an emerging technology, which is capable of acquiring dynamic data related to a human body's physiological conditions. The collected data can be applied to detect anomalies in a patient, so that he or she can receive an early alert about the adverse trend of the health condition, and doctors can take preventive actions accordingly. In this paper, a new WWSN for anomaly detections of health conditions has been proposed, system architecture and network has been discussed, the detecting model has been established and a set of algorithms have been developed to support the operation of the WWSN. The novelty of the detected model lies in its relevance to chronobiology. Anomalies of health conditions are contextual and assessed not only based on the time and spatial correlation of the collected data, but also based on mutual relations of the data streams from different sources of sensors. A new algorithm is proposed to identify anomalies using the following procedure: (1) collected raw data is preprocessed and transferred into a set of directed graphs to represent the correlations of data streams from different sensors; (2) the directed graphs are further analyzed to identify dissimilarities and frequency patterns; (3) health conditions are quantified by a coefficient number, which depends on the identified dissimilarities and patterns. The effectiveness and reliability of the proposed WWSN has been validated by experiments in detecting health anomalies including tachycardia, arrhythmia and myocardial infarction.